Update README.md

README.md

@@ -20,7 +20,7 @@ p{color:Black !important;}
 
 
 <p align="center">
-      <img src="https://ibm.box.com/shared/static/7vq37r4ny12whwcw7les0txr3l2ltdys.png" width=70%/>
+      <img src="https://github.com/javiccano/Wikipedia_contradict_benchmark/blob/main/figs/Example.png?raw=true" width=70%/>
 <!--  <img src="./figs/Example.png" width=70%/> -->
 </p>
 
@@ -67,14 +67,14 @@ The dataset has been used in the paper to assess LLMs performance when augmented
 The following figure illustrates the evaluation process:
 
 <p align="center">
-  <img src="https://ibm.box.com/s/xstzgegdlvqejvd2q12pld1c82vcbhs9" width=70%/>
+  <img src="https://github.com/javiccano/Wikipedia_contradict_benchmark/blob/main/figs/Evaluation.png?raw=true" width=70%/>
 <!--  <img src="./figs/Evaluation.png" width=70%/> -->
 </p>
 
 And the following table shows the performance of five LLMs (Mistral-7b-inst, Mixtral-8x7b-inst, Llama-2-70b-chat, Llama-3-70b-inst, and GPT-4) on the Wikipedia Contradict Benchmark based on rigorous human evaluations on a subset of answers for 55 instances, which corresponds to 1,375 LLM responses in total.
 
 <p align="center">
-  <img src="https://ibm.box.com/shared/static/brqfm2g2q5g63cxirsaa5jf0apqmbe5d.png" width=70%/> 
+  <img src="https://github.com/javiccano/Wikipedia_contradict_benchmark/blob/main/figs/table2.png?raw=true" width=70%/> 
   <!-- <img src="./figs/table2.png" width=70%/> -->
 </p>